Known communication systems utilize data conversion from voice data to text data or data conversion from text data to voice data.
A technique disclosed in Patent Literature 1, for example, prepares in advance first dictionary information (“word dictionary”) that is used for converting text data into voice data. The technique converts input text data into a corresponding phonetic character string using the first dictionary information and processes the converted phonetic character string on the basis of predetermined information included in the input text data so as to obtain synthesized voice to which, for example, feeling, thought, sex, looks, body height, body weight, age, occupation, hometown, hobby, and taste are added.
The technique disclosed in Patent Literature 1 further includes second dictionary information (“word model” and “language model”) prepared in advance and used for converting voice data into text data. The technique converts input voice data into text data using the second dictionary information, estimates from the input voice data, feeling, thought, physical condition, sex, looks, body height, body weight, age, occupation, hometown, hobby, and taste, for example, of a speaker, and processes the converted text data in accordance with a result of the estimation.